Voltar aos Detalhes do Artigo From 'black box' to 'glass box': using Explainable Artificial Intelligence (XAI) to reduce opacity and address bias in algorithmic models Baixar Baixar PDF